Page 95 - Kaleidoscope Academic Conference Proceedings 2022

Extended reality – How to boost quality of experience and interoperability

on the architectures presented in the previous sections, are then see a list of all building units contained within the
now all published in the KKG and made available to the AR building in order to choose which unit in particular they
application based on an endpoint. would like to see information on. This functionality in the
application is illustrated in Figure 5 and is supported by the
For the application to provide users with the information second query which takes the building identifier from the
seen in the figure above, the KKG endpoint needs to be first query as input and returns all information about building
queried by three different queries and that data made units contained within that building, including house
available through an API to be integrated into the application numbers, letters and additions as well as the postcode and
itself. Both queries are SPARQL queries being performed on street name associated with that address. Naturally, this
the SPARQL endpoint for the KKG and the resulting APIs functionality is most useful when multiple units exist within
for each query are simply integrated as API variables in the a building but is still available in the application when only
application. In line with the system architecture illustrated in a single unit is present. The selection made in this step is used
Figure 3, SPARQL APIs are used as the interface between as input for the following step.
the KKG as the application-independent data source and the
AR application itself. As previously noted, SPARQL, an
open standard, is the most widely used query language for
linked data and allows users to query data flexibly and make
use of the returned data in an application given that the user
has knowledge about the structure of the data model. The
linked data available through a SPARQL API can be
5
returned in a range of formats including JSON-LD, Turtle
6
and N-Quads . These formats are native linked data
serialization formats, each with a different structure.
The first query, which is available here, has a single input
query parameter in the form of a point geometry. This
geometry is defined by the application itself in fetching the
latitude and longitude of the user’s location, combining these
to form the point geometry string and using this string as
input for the first SPARQL query. The query returns the
building identifier and associated polygon for each building
within a 100 meter radius of the defined point (i.e. the user’s
location) defined by the application. The results of this query
are visualized only as a house icon above a building (see
Figure 4) where the icon is centered on a building based on
the central point of the polygon returned during this query.
This first query is introduced to support better performance Figure 5 – Screenshots displaying building units from
in the application. Indeed, querying building information the augmented reality application built on the
within a radius requires a large amount of computational Kadaster Knowledge Graph
power and this query returns a minimal amount of
information which then allows successive queries to be The final query, which is available here, makes use of the
performed on a subset of the data, improving overall identifier associated with the building unit selected by the
performance of the application. user and then displays all information associated with both
the whole building such as the building year and statistical
Once the house icon has been visualized in the application, information, as well as the information only associated with
the user is able to tap on the house icon to indicate that they a given unit, including floor size and usage function. An
are interested in seeing more information about that example of this information is visualized and presented to
particular building. In selecting a given building, the the end user and can be seen in Figure 4.
building identifier associated with that icon is then used as
the input parameter for the second query, which is available As more information becomes available in the KKG from
here. The KKG data model makes a distinction between new data sources, the same endpoints can be used to display
buildings and building units (in Dutch: verblijfsobjecten), more detailed information about a building in the AR
where the attributes such as building year are associated with application. The only changes that would need to be made
a whole building and usage function and floor size are are in the application itself to transform and render the
associated with the building unit. In the application, this information for display on the screen in a similar manner as
distinction allows the users to click on the house icon and above. From the user perspective, the AR application serves

5 https://www.w3.org/TR/turtle/ 6 https://www.w3.org/TR/n-quads/ .

– 49 –